1 Overview of GAGA assemblies

1.1 Taxon sampling

We so far have genomic data for 163 species distributed across the 347 genera of ants (Fig 1.1), covering 27 tribes in 12 subfamilies. Myrmicinae are represented by 77 genomes, followed by Formicinae and Ponerinae with 39 and 21 genomes sequenced, respectively. With regard to the tribe, Crematogastrini make up the largest group with 35 genomes, followed by Attini and Ponerini with 25 and 20 genomes sequenced, respectively.

**Ant genera targeted so far by GAGA.** The Figure shows all extant genera of the Formicidae with the corresponding species sequenced by GAGA for the different genera.

Figure 1.1: Ant genera targeted so far by GAGA. The Figure shows all extant genera of the Formicidae with the corresponding species sequenced by GAGA for the different genera.

1.2 Sequencing techniques

Most (n= 125) of the species have been sequenced with PacBio, while some rare species (e.g. Leptanilla) could only be sequenced with stLFR (n= 20), commonly due to lack of sufficient biomass (Fig. 1.2). HiC sequencing was so far done for 15 species, allowing us to assemble chromosome-scale genomes for 18 species. Species highlighted in grey in the phylogeny were not assembled by GAGA but included in our comparative analyses as they met our quality criteria.

Most genomes reach high BUSCO scores (Fig ??), both against Hymenoptera (median 97 %) and Eukaryota (median 98 %). A few stLFR-only genomes however are incomplete, with BUSCO scores as low as ~70 % (e.g. Gigantiops destructor with 68.3 % complete Hymenoptera BUSCOs).

We have now also finalized the gene annotations for all species. On average, we annotated 13182 genes (range 10973 to 16903). BUSCO scores for the protein annotations were highly correlated with scores for the unannotated genome, indicating that our gene annotation approach is working well.

Sequencing techniques for different species. The left panel shows the techniques combined for genome sequencing of the different species. The right panel shows the coverage of RNAseq data, divided by caste, sex, and developmental stage.

Figure 1.2: Sequencing techniques for different species. The left panel shows the techniques combined for genome sequencing of the different species. The right panel shows the coverage of RNAseq data, divided by caste, sex, and developmental stage.

1.3 Genome sizes of all so far sequenced species

The assembled genomes vary considerably in size, ranging from 188.61 MB (Cardiocondyla obscurior) to 594.51 MB (Odontomachus cf. monticola). Overall, genome size tends to vary most in the Ponerinae and least in the Formicinae. Figure 1.3 gives a complete overview of all so far sequenced species and the respective genome assembly size in Mb.

Figure 1.3: Genome sizes of GAGA genome assemblies

1.4 Preliminary quality of the assembled genomes

Our sequencing strategy has evolved further, now combining long-read PacBio data, linked-short-read stLFR data and if possible HiC data. We expect that together, these data will yield highly contiguous high quality genome assemblies that should be close to chromosome-resolution for most of the species.

Figure 1.4 gives a rough overview of the quality of the genome assemblies we have so far.

On average, scaffold N50 is 4.55 mega bases (Mb) Currently, the species with the highest scaffold N50 is Aphaenogaster subterranea with 35.8 MB.

Figure 1.4: Genome assembly quality for GAGA species.

Figure 1.5: Genome assembly quality overview for GAGA species.

1.5 Overview Table

Table 1.1 gives an overview of all so far sequenced species.